Skip to main content

Advertisement

Table 3 Fixed effects of general and generalized linear models for infection levels in eight Tropheus populations

From: Contrasting parasite communities among allopatric colour morphs of the Lake Tanganyika cichlid Tropheus

     Infection presence Abundance Infection intensity
Parasite group Effect Num DF Den DF F P F P F P
Acanthocephalans site 7 605/483 9.01 < 0.0001 23.46 < 0.0001 11.92 < 0.0001
  day 2 605/483 0.76 0.4673 1.02 0.3622 01.25 0.2886
  sex 1 605/483 0.01 0.9100 0.04 0.8484 0.04 0.8351
  SL 1 605/483 12.62 0.0004 57.06 < 0.0001 47.92 < 0.0001
  year(site) 6 605/483 5.02 < 0.0001 7.57 < 0.0001 5.02 < 0.0001
Nematodes site 7 605/47 2.38 0.0210 2.64 0.0106 1.32 0.2640
  day 2 605/47 0.63 0.5336 0.78 0.4591 2.41 0.1005
  sex 1 605/47 1.09 0.2974 1.36 0.2434 1.17 0.2843
  SL 1 605/47 0.10 0.7557 0.05 0.8185 1.39 0.2436
  year(site) 6 605/47 0.89 0.5011 2.19 0.0428 0.74 0.6009
Gyrodactylus site 7 616/24 3.95 0.0003 5.47 < 0.0001 0.83 0.5692
  day 2 616/24 0.90 0.4091 0.37 0.6927 4.88 0.0167
  sex 1 616/24 0.30 0.5811 0.09 0.7607 1.16 0.2926
  SL 1 616/24 0.26 0.6093 0.27 0.6016 1.25 0.2751
. year(site) 6 616/24 0.35 0.9071 0.60 0.7344 0.93 0.4933
Cichlidogyrus site 7 476/418 2.85 0.0064 18.85 < 0.0001 15.14 < 0.0001
  day 2 476/418 2.80 0.0617 5.50 0.0043 2.54 0.0803
  sex 1 476/418 3.62 0.0578 7.58 0.0061 4.00 0.0462
  SL 1 476/418 6.60 0.0105 77.90 < 0.0001 74.38 < 0.0001
  year(site) 6 476/418 2.80 0.0108 3.68 0.0014 1.73 0.1131
Ergasilus site 7 476/158 1.73 0.1001 2.40 0.0201 1.86 0.0792
  day 2 476/158 0.70 0.4984 0.14 0.8729 1.36 0.2602
  sex 1 476/158 0.15 0.6978 0.10 0.7479 0.00 0.9537
  SL 1 476/158 0.58 0.4485 3.28 0.0708 5.30 0.0227
  year(site) 6 476/158 0.70 0.6489 0.55 0.7696 0.23 0.9667
  1. Fixed effects included sampling site, processing day, sex, standard length (SL), and sampling year (nested in site). Observer effects were included as random (not shown), except for Gyrodactylus sp. for which there was only one observer. The model for infection presence assumes a binomial distribution and models the logit of the probability of infection. For the models for abundance and infection intensity, the dependent variable was square-root transformed. Note that the denominator degrees of freedom (Den DF) are higher for the infection presence and abundance model (before the dash) than for the infection intensity model (after the dash). P-values in bold indicate significance at α = 0.05.